Semiconductor power devices have been in use since the early 1950s. They are specialized devices used as switches or rectifiers in power electronics circuits. Semiconductor power devices are characterized by their ability to withstand high voltages and large currents as well as the high temperatures associated with high power operation. For example, a switching voltage regulator will comprise two power devices that constantly switch on and off in a synchronized manner to regulate a voltage. The power devices in this situation need to sink system-level current in the on state, withstand the full potential of the power supply in the off state, and dissipate a large amount of heat. The ideal power device is able to operate in high power conditions, can rapidly switch between on and off states, and exhibits low thermal resistance.
A standard power device structure implemented using Metal-Oxide Semiconductor Field Effect Transistor (MOSFET) technology is the Vertical Diffused Metal-Oxide Semiconductor (VDMOS) structure. The VDMOS structure is also known as Double-diffused MOS (DMOS). The “vertical” term is used because current flows vertically through the device, and the “diffused” term is used because the channel and source regions are produced through a diffusion processing step. The structure can be described with reference to FIG. 1.
FIG. 1 displays a cross-section of a VDMOS power device 100. The power device 100 includes one or more source electrodes 101, a drain electrode 102, and a gate electrode 103. Source regions 104 are N+ doped in an n-type VDMOS device. In contrast to a standard MOSFET configuration, the source regions 104 are located on either side of a gate 105 below a gate insulator 106. Channel regions 107 are P+ doped in an n-type VDMOS device, and they are disposed between a drain region 108 and the source regions 104. In an n-type VDMOS device a high voltage applied to the gate electrode 103 will invert the channel regions 107 between the source regions 104 and the drain region 108. This configuration allows the power device 100 to withstand both a high voltage in the off state and a high current in the on state as compared to a standard MOSFET implemented using the same amount of die area. The channel width of the power device 100 is double that of a traditional MOSFET with the same die area thereby allowing the power device 100 to withstand large currents. In addition, the dimension that would usually be the channel length in a traditional MOSFET does not affect the breakdown voltage. Instead, the thickness and doping of the drain region 108 determines the breakdown voltage of the power device 100. The drain region 108 is usually the device substrate when a VDMOS device is implemented in a regular bulk semiconductor process.
The VDMOS power device 100 has certain disadvantageous aspects that limit it from performing as an ideal power device. For instance, there is a large junction capacitance formed by the boundary between the drain region 108 and the channel region 107. This capacitance is generally due to an area component set by a dimension 111 and a depth component set by a dimension 110. Since the junction formed by the drain region 108 and the channel region 107 must be charged or discharged when the power device 100 switches state, the capacitance of this junction degrades the performance of the power device 100. In addition, since the area component is limited, it is not possible to contact the source regions 104 and the channel regions 107 separately, since electrodes such as source electrode 101 can often consume a large amount of area. Furthermore, the power device 100 suffers from very poor thermal performance, since it is implemented on bulk semiconductor. Power devices implemented in bulk semiconductor typically have a minimum wafer thickness of approximately 200 μm due to the high incidence of wafer breakage when handling large-diameter wafers thinner than that. Since the thermal resistance of a silicon substrate is proportional to the thickness of the silicon substrate, the implementation of power devices on bulk semiconductor is problematic in terms of thermal performance. A high level of heat in an integrated circuit can shift the electrical characteristics of its devices outside an expected range causing critical design failures. Left unchecked, excess heat in a device can lead to permanent and critical failures in the form of warping or melting materials in the device's circuitry.
Additionally, layer transfer technology typically involves a pair of semiconductor wafers at various stages of processing that are bonded together using direct, molecular, or adhesive bonding. If one of the wafers is a semiconductor-on-insulator (SOI) or silicon-on-insulator wafer with the substrate removed to expose the buried oxide, the resulting structure comprises a device layer that is upside-down with respect to its original orientation and that has been transferred from an SOI wafer to a new handle wafer.
A layer transfer structure 200 is shown in FIG. 2. The layer transfer structure 200 includes a handle wafer 201 and an SOI wafer 202. The handle wafer 201 comprises a handle wafer substrate 203 and a handle bond layer 204. The SOI wafer 202 comprises an insulator layer 205 and a circuitry layer 206. The layer transfer structure 200 illustrates the finished product of a layer transfer process. However, before layer transfer begins, the SOI wafer 202 additionally comprises another layer of substrate material below the insulator layer 205. The substrate layer is typically a semiconductor material such as silicon. The insulator layer 205 is a dielectric which is often silicon-dioxide formed through the oxidation of the substrate silicon. The circuitry layer 206 includes a combination of dopants, dielectrics, polysilicon, metal layers, passivation, and other layers that are present after structures 207 have been formed therein. The structures 207 may include metal wiring; passive devices such as resistors, capacitors, and inductors; and active devices such as transistors. Layer transfer begins when the handle bond layer 204 is bonded to the top of the SOI wafer 202. At this point, the handle wafer 201 provides sufficient stability to the SOI wafer 202 such that the aforementioned layer of substrate material below the insulator layer 205 can be removed. As a result of this process, the layer transfer structure 200 provides a device that can be contacted through a bottom surface 208. This means that external contacts to the structures 207 in the circuitry layer 206 are extremely close to the structures 207 themselves. In some situations this distance is on the order of 1 micro-meter (μm).
As used herein and in the appended claims, the “top” of the layer transfer structure 200 references a top surface 209 while the “bottom” of the layer transfer structure 200 references the bottom surface 208. This orientation scheme persists regardless of the relative orientation of the circuitry layer 206 to other frames of reference, and the removal of layers from, or the addition of layers to the SOI wafer 202. Therefore, the circuitry layer 206 is always “above” the insulator layer 205. In addition, a vector originating in the center of the circuitry layer 206 and extending towards the bottom surface 208 will always point in the direction of the “back side” of the layer transfer structure regardless of the relative orientation of the SOI wafer 202 to other frames of references, and the removal of layers from, or the addition of layers to the SOI wafer 202.
It is with respect to these and other background considerations that the present invention has evolved.